DESCRIPTION: Based on the argument that "behavioral observation by coders blind to group assignment appears to be the most sensitive and accurate way of measuring intervention impact," the applicant proposes three ways of refining existing observational techniques to enhance their accuracy and usefulness. First, the applicant will refine and further test a low-cost paper and pencil observational coding system. Second, the applicant will develop ways of allowing coders to code one person at a time or to apply multiple coding frameworks to the same data and to synchronize and merge the resulting streams of data. Using digital media will make it much faster to retrieve and access particular behavioral interactions for coding and analysis. Third, the applicant will develop statistical and graphical methods for summarizing observational data, and integrate these so that researchers can examine real-time video, behavioral codes, and statistical and graphical summaries of the same data, as a tool for theory development. For the second two projects, staff will attempt to use and modify commercially available software to make their systems more readily usable by others.